OpenAI Integrates Codex Programming Model into GPT-5.5 Architecture

OpenAI has recently announced a major technological integration: the dedicated Codex programming model will no longer be offered as a separate product. Its core capabilities are now fully integrated into the GPT-5.5 main model, meaning that after the GPT-5.4 version, there will be no dedicated programming branch on the market, and GPT-5.3 serves as the final standalone Codex release.
This shift reflects a deeper evolution in OpenAI's R&D approach, moving from a specialized plug-in to an inherently versatile model. Developers used to rely on Codex, which was specifically optimized for code. However, given the significant performance gains in general large language models, OpenAI now believes that embedding programming capabilities as a core attribute within the main model allows for more efficient resource integration. In practice, the integrated GPT-5.5 shows notably better resource utilization when handling complex programming tasks.
For developers, the most immediate change is in cost and efficiency. While the API call price for GPT-5.5 has increased by about 20%, the model's improved comprehension means the number of tokens consumed for similar programming tasks has actually gone down. This "higher unit price, lower consumption" dynamic highlights the model's more accurate and efficient processing of logical code.
Looking back at Codex's history, it experienced cycles of being closed and then reintroduced. Its full integration into the main model now signals that AI programming is no longer a special skill requiring separate maintenance, but a foundational metric for measuring the intelligence of general large models. OpenAI's move not only streamlines its product line but also sends a clear message: future AI will no longer rely on fragmented toolkits—a sufficiently powerful general intelligence should inherently possess top-tier engineering capabilities.
Related article
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
How Recursive Intelligence Raised $335M at $4B Valuation in 4 Months
The co-founders of startup Ricursive Intelligence seemed destined to be co-founders.Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got those weird emails from Zuckerber
Related Special Topic Recommendations
Comments (0)
0/500

OpenAI has recently announced a major technological integration: the dedicated Codex programming model will no longer be offered as a separate product. Its core capabilities are now fully integrated into the GPT-5.5 main model, meaning that after the GPT-5.4 version, there will be no dedicated programming branch on the market, and GPT-5.3 serves as the final standalone Codex release.
This shift reflects a deeper evolution in OpenAI's R&D approach, moving from a specialized plug-in to an inherently versatile model. Developers used to rely on Codex, which was specifically optimized for code. However, given the significant performance gains in general large language models, OpenAI now believes that embedding programming capabilities as a core attribute within the main model allows for more efficient resource integration. In practice, the integrated GPT-5.5 shows notably better resource utilization when handling complex programming tasks.
For developers, the most immediate change is in cost and efficiency. While the API call price for GPT-5.5 has increased by about 20%, the model's improved comprehension means the number of tokens consumed for similar programming tasks has actually gone down. This "higher unit price, lower consumption" dynamic highlights the model's more accurate and efficient processing of logical code.
Looking back at Codex's history, it experienced cycles of being closed and then reintroduced. Its full integration into the main model now signals that AI programming is no longer a special skill requiring separate maintenance, but a foundational metric for measuring the intelligence of general large models. OpenAI's move not only streamlines its product line but also sends a clear message: future AI will no longer rely on fragmented toolkits—a sufficiently powerful general intelligence should inherently possess top-tier engineering capabilities.
Stripe to Acquire OpenRouter as AI Model Routing Gains Momentum
Stripe has agreed to acquire OpenRouter, an AI model-routing platform that gives developers access to hundreds of models through a single interface. The deal adds model selection and routing to Stripe’s existing work around AI usage and token-based b
Former OpenAI Researcher Barrett Zoph Returns to Google to Support Gemini Development
Barret Zoph, who previously abruptly left and joined OpenAI due to a disagreement with Thinking Machines Lab CEO Miriam Mularati, has once again changed employers. Alphabet's Google announced that Zoph will return as Vice President of Research, bring
How Recursive Intelligence Raised $335M at $4B Valuation in 4 Months
The co-founders of startup Ricursive Intelligence seemed destined to be co-founders.Anna Goldie, CEO, and Azalia Mirhoseini, CTO, are so well-known in the AI community that they were among those AI engineers who “got those weird emails from Zuckerber





Home






